Species richness and cover along a 60-year chronosequence in old-fields ofsoutheastern Spain

Andreu Bonet1,* and Juli G. Pausas1,2

1Department of Ecology, Faculty of Sciences, University of Alicante (UA), A.C. 99, E-03080 Alicante, Spain;2Centro de Estudios Ambientales de Mediterráneo (CEAM), Charles R. Darwin 14, Parc Tecnològic, 46980Paterna, València, Spain; *Author for correspondence (e-mail: andreu@ua.es)

Received 25 June 2002; accepted in revised form 19 September 2003

Key words: Abandonment, Dispersal, Life form, Old-fields, Mediterranean, Non-linear trends, Semi-arid, Spe-cies richness, Succession

Abstract

We analyse changes in plant cover and species richness along a 60-year chronosequence in semi-arid Mediter-ranean old-fields of southeastern Spain. The objectives were: �i� to study patterns of species richness along theabandonment gradient in semi-arid conditions �e.g., to test the “humped-back model” in our system�; �ii� to testwhether different broad life forms �annuals, forbs, grasses and woody species� showed different patterns alongthe abandonment gradient, and �iii� to examine to what extent plants with different dispersal strategies dominateat different stages of succession. The explained variance of the regression relating species richness to years sinceabandonment is improved when considering different life forms. The results suggest that cover and richness ofdifferent functional groups show a non-linear unimodal �often positive-skewed� pattern along the gradient �agesince abandonment�. Maximum total richness is found at young stages of abandonment � � 20 years�, when mostlife forms and dispersal strategies coexist. Annuals and perennial forbs reached their maximum richness duringthe first 10 years of abandonment. About 45% of total woody species richness is reached at this time as a con-sequence of early colonization of zoochorous shrubs. While the results showed a tendency towards a life-formreplacement sequence, the pattern is not so clear when looking at the different dispersal strategies. The resultscomplement previous results in Mediterranean conditions and emphasise the importance of considering differentfunctional types when studying successional patterns.

Introduction

Mediterranean landscapes of Europe are sufferingfrom the rural exodus caused by the socioeconomicchanges of the last decades. These changes in tradi-tional land-uses and lifestyles have resulted in theabandonment of large areas of farm-land �Lepart andDebussche 1992; Lasanta and García-Ruiz 1996�. Theresulting build-up of early successional vegetationhas strong implications on water balances �Bellot etal. 2001� and fire regimes �Rego 1992; Pausas andVallejo 1999; Pausas in press�.

Secondary succession in Mediterranean ecosystemshas been studied extensively in mesic conditions�Houssard et al. 1980; Escarré et al. 1983; Peco et al.1991; Lavorel et al. 1994; Tatoni and Roche 1994;Tatoni et al. 1994; Montalvo et al. 1995; Debusscheet al. 1996; Ne’eman and Izhaki 1996; Bonet 1997�,but few studies have been made in semi-arid environ-ments �e.g., Noy-Meir 1973; Martínez-Fernández1995; Margaris et al. 1996; Bonet et al. 2001; Bonet,2004�. In dry environmental conditions, vegetationcover tends to be low and sparse �Schlesinger et al.1990�, and interspecific interactions such as competi-tion or facilitation may be different from those in

Plant Ecology 174: 257–270, 2004.© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

257

more mesic Mediterranean conditions �e.g., Pugnaireand Luque 2001; Maestre et al. 2003�. Under aridconditions, two vegetation parameters are particularlyrelevant: plant cover �as a factor for soil protectionand for enhancing ecosystem processes� and richness�a measure of biodiversity�. In the present work, weattempt to study these two parameters during a 60-year chronosequence in semi-arid old-fields of thesoutheastern Iberian Peninsula.

During the early stages of succession, the gradualcolonisation and increase in spatial heterogeneityshould lead to an increase in species richness. Subse-quently, species richness may drop as a result ofcompetitive interactions �Bazzaz 1975; Huston 1979;Tilman 1982; Peet and Christensen 1988�. Thus, theso-called “humped-back model” �unimodal responsecurve� of species richness proposed for nutrient andproductive gradients �Grime 1973; Wheeler andGiller 1982; Tilman 1982; Wisheu and Keddy 1989;see the review by Pausas and Austin 2001� and dis-turbance frequency gradients �Connell 1978; Lub-chenco 1978; Huston 1979; Wilson and Keddy 1988�could also be applied to the time since abandonment�e.g., Peet 1978�, although the underlying factors maybe different in each of these cases.

However, some successional studies in Mediterra-nean climates have found a decreasing �Debussche etal. 1996� or fluctuating pattern �Houssard et al. 1980�of species richness after abandonment, and Noy-Meir�1973� found little changes in diversity through timein arid conditions. Furthermore, several studies sug-gested that after perturbation, the pattern of richnessshould be different in different environmental condi-tions �Auclair and Goff 1971; Peet 1978; Huston1994�. These researchers suggested that under mesicconditions species richness will show a peak at earlystages of succession, while on sites under stressfulconditions �e.g., xeric sites� a steady increase in di-versity is expected.

Because species of the same functional type sharemany characteristics, competitive interactions amongspecies of the same functional types may be strongerthan between species of different functional types,and so, patterns of species richness along gradientsmay be more interpretable by considering the speciesrichness of the different functional types �Peet 1978;Pausas 1994; Pausas and Austin 2001�. In this workwe consider two functional classifications of species,one based on life form and the other on dispersalstrategy. Life forms are well-known broad functionaltypes related to longevity and resource acquisition

�e.g., Wright 1992; Lavorel et al. 1997�. If successionfrom abandonment is determined by the ability ofplant arrival �colonisation capacity; van der Valk1992; Willson 1993�, then, plants with different dis-persal strategies should have differential dominancealong succession �i.e., the ‘relay floristics’ model;Egler 1954; Myster 1993�. The dispersal mechanismadopted by species is a key trait influencing thestructure of communities �Brown 1992�.

Our aims were: 1� to analyse the changes in plantcover and species richness along a land abandonmentgradient in a semi-arid ecosystem; 2� to test whetherdifferent broad life forms �annuals, forbs, grasses andwoody species� showed different patterns along thetemporal gradient; 3� to test to what extent plants withdifferent dispersal strategies dominate at differentstages of land abandonment.

To answer these questions, we used the time sinceabandonment as an indicator of the successional gra-dient. For practical reasons, many of the Mediterra-nean secondary succession studies in old-fields haveused a chronosequence approach instead of diachro-nic studies �Escarré et al. 1983; Tatoni and Roche1994; Bonet 1997�. Despite the existence of severalmethodological problems using chronosequencing�Lepart and Escarré 1983; Glenn-Lewin and van derMaarel 1992�, most of the predictions made withsome of these synchronic studies or space-for-timesubstitutions �sensu Pickett 1989� have been validatedby revisiting and re-sampling the studied communi-ties �Debussche et al. 1996; Foster and Tilman 2000�.

Methods

Study area

The study was conducted at the Ventós-Agost Catch-ment Experimental Station �University of Alacant�, inthe Municipality of Agost, Alacant Province, SESpain �38° 28'N, 0° 37'W, 10-840 m a.s.l.�. Thecatchment area �approx. 1537 ha� is characterised bya semi-arid Mediterranean climate, with a very highinterannual variability. Mean annual temperature is18.2 ºC and annual rainfall is 302.1 mm, most ofwhich falls in autumn �Agost Meteorological Station,1961-1990 period�. Soils have developed over marlsand calcareous bedrock, and slopes vary between 25-30%, and are mainly south-facing. Other soil charac-teristics are summarised in Bellot et al. �1999�.

258

The present vegetation mosaic of the overall areais dominated by Stipa tenacissima L. steppes, result-ing from the disturbance of shrublands �Rivas-Mar-tínez 1987�, but probably also resulting from aban-doned Stipa plantations in non-arable lands �Barberet al. 1997�. Other important communities are drygrassland formations of Brachypodium retusum Pers.Beauv. with dwarf scrubs, and scattered shrublandswith Quercus coccifera L.

The lower part of the catchment is occupied bytraditional agriculture terraces currently abandonedand surrounded by steppes and the described vegeta-tion mosaic. Field abandonment ranged from 1 to 60years following the last cultivation, but the main landabandonment process was experienced during the pe-riod 1946-1956 �Bonet et al. 2001�. Abandoned fieldson terraces were mainly composed of tree crops, suchas almond �Prunus dulcis �Miller� D.A. Webb�, olive�Olea europaea L.� and carob trees �Ceratonia sili-qua L.�. Traditional agricultural practices such as till-ing and irrigation regimes were quite homogeneousfor all these tree crop stands before abandonment, butthere were also some scattered cereals and horticul-ture abandoned crops in the area. The great speciescomposition variability associated with differentmanagement and land uses of the fields �Bonet, inpress� justifies the use of life forms rather than spe-cies in the trend analysis along the analysed gradient.

Sampling

Vegetation was sampled through a chronosequence ofold-fields. During spring 1999 and 2000, we sampled96 plots of abandoned croplands on field terraces. Theabandonment is defined here as the cessation ofploughing and sowing, but other uses and activitiessuch as grazing could still be present in the fields�Bonet, 2004�.

We first estimated the period of abandonment byusing a sequence of aerial photographs �1946, 1956,1974, 1980, and 1995�. Then we consulted the AgostMunicipal Archives and held personal interviews withlandowners and managers in order to determine theage of abandonment more precisely.

On each old-field terrace, a 10 � 10 m �100 m2�plot was delineated avoiding edges and field margins�i.e., with at least a 2 m-wide buffer zone�. For thewhole 10 � 10 m-plot a complete species list was re-corded and total species richness was calculated. Ineach of these plots, 20 quadrats of 1 � 1 m weremarked off in a systematic sampling procedure along

four transects �5 quadrats per transect�. Each transectwas separated from its neighbour by 2 m, and thedistance between quadrats in each transect was 1 m.The species cover �%� was visually estimated in eachquadrat and then an average cover was calculated forall the quadrats in a plot �total sampled area was 20m2�.

Data analysis

Species were grouped by life forms �annuals or bien-nials, perennial forbs, perennial grasses and woodyspecies� and by their main dispersal strategy �anemo-chory, barochory, ectozoochory, endozoochory, andmirmecochory� �see Appendix�. The main dispersalmode of each species was based on field observationsand following Molinier and Müller �1938�, van derPijl �1972� and Bonet �1997�; secondary dispersalstrategies were not considered in this study. Coverand richness were computed for all species as well asfor each life form and each dispersal mode. Coverwas calculated on each old-field from the sampledquadrats, and richness was determined by the numberof species in the whole 10 m � 10 m plot.

Cover and richness patterns �total, for the differentlife forms and for the different dispersal strategies�were related to years since abandonment by leastsquare regressions using SPSS �v. 11.0� statisticalsoftware. Three different types of regression equa-tions were tested �Linear, Gaussian and Lognormal�and the most significant was chosen. In most cases themost significant was the non-linear lognormal regres-sion because it allowed us to detect skewed unimodalpatterns �Draper and Smith 1981; Scales 1985; Sokaland Rohlf 1995�. The goodness of fit was evaluatedusing the F-test. Parameters were tested by a t-testand their standard errors were used as a gauge of theaccuracy of the fitted curve. We also searched for pa-rameter dependencies in order to avoid over-param-eterisation of the regression equations.

Results

A total of 222 species were recorded in the wholestudy area �see Appendix�. Total plant cover rangedfrom 45.5 to 100% �Table 1�, and this variation wasnot related to age since abandonment �p � 0.4189 forthe lognormal model�. Some recently abandoned plotshad close to 100% cover. Species richness per plot�100 m2� ranged from 4 to 34, and there was a sig-

259

nificant non-linear relationship with age �R2 � 0.36,p � 0.0001; Figure 1�.

Life forms

Of the 222 species, the most represented life formwas annual species �36%�. Forbs and woody specieswere 29 and 27% respectively, and grasses were rep-resented by 8% of the species. At plot scale �i.e., 10

� 10 m�, woody species were the most abundantspecies �mean woody species richness � 7 / 100 m2�,followed by forbs �6.3�, annuals �4.4� and thengrasses �1.7� �Table 1�. Mean cover of the differentlife forms ranged from 22% �annuals� to 26% �woodyspecies�. However, both species richness and covershowed great variation �Table 1�, which can be ex-plained by plot age �time since abandonment; Figure1, Figure 2, Figure 3; Table 2�.

Table 1. Summary of cover �%� and species richness �per 100 m2� in all old-fields for the different life forms and dispersal types considered.

Cover �%� Species richness

mean sd min max Mean sd min max

Life formAnnuals 22.18 29.64 0.00 100 4.40 4.33 0 17Grasses 27.54 22.58 0.00 75.76 2.43 1.65 0 8Forbs 23.46 20.88 0.00 82.61 6.33 3.31 0 16Woodies 26.45 25.73 0.00 93.68 6.97 5.54 0 21

DispersalAnemochorous 45.40 21.03 1.00 93.10 7.86 3.56 1 18Mirmecochorous 1.35 3.88 0.00 22.50 0.40 0.59 0 2Barochorous 37.35 22.80 3.75 90.20 9.13 4.20 2 19Ectozoochorous 11.92 16.80 0.00 61.11 1.94 1.62 0 6Endozoochorous 53.58 8.67 0.00 59.20 0.80 0.99 0 4

Total 96.90 9.72 45.50 100 19.76 6.41 4 34

Figure 1. Total species richness �on 100 m2-plots� along the successional gradient. Fitted line is the significant non-linear regression �seeTable 2 for more details�.

260

Figure 2. Plant cover �%� of different life forms along successionalgradient. Fitted lines are the significant non-linear and linear re-gressions �see Table 2 for more details�.

Figure 3. Species richness �on 100 m2-plots� of different life formsalong successional gradient. Fitted lines are the significant non-lin-ear regressions �see Table 2 for more details�.

261

The cover of each life form showed a significantand different pattern along the abandonment age gra-dient, and there was a clear tendency in the order ofcover dominance of the different life forms �Figure 2,Table 2�. The explained variance of the regressionsranged from 0.42 �forbs� to 0.89 �woody species�.Annuals, forbs and grasses showed a skewed patternwhile woody species showed a linear tendency. An-nuals were the first species covering the old-fields andreached up to 100% of the soil in the first years.However, they were almost absent ca. 12 years afterabandonment. Perennial forbs also reached theirmaximum during the first 10 years, while perennialgrasses peaked at 10-25 years after abandonment.Woody species showed a significant monotonicincrease throughout the time window studied �60years�.

Species richness also showed different patterns forthe different life forms �Figure 3, Table 2�, and theorder along the age gradient was similar to cover; themain difference was that there were few species ofperennial grasses. The explained variance in the log-normal regressions was higher in all cases than the

explained variance for total species richness �0.36�,and ranged from 0.38 �grasses� to 0.67 �annuals�.

The number of life forms decreased with abandon-ment age, from 4 co-occurring during the first 20years, to 2 life forms, and finally �at ~ 60 years� thevegetation was practically dominated by one life form�woody species�, although one grass species �Brachy-podium retusum� persisted �with low cover� on the60-year-old plots.

Dispersal strategy

Of all the species recorded in the study area, mostwere dispersed by gravity �43%� or by wind �42%�; afew were considered ectozoochorous �9%�, endozoo-chorous �3%� and mirmecochorous �3%� species. Asimilar ranking was observed for mean species rich-ness at the plot level �Table 1�. Mean plant cover washighest for endozochory �54%� and anemochory�45%�, intermediate for barochory �37%� and rela-tively low for ectozoochory �12%� and mirmecochory�1%�. In all cases, the variability of these values wasvery large �Table 1�. The cover pattern for each dis-persal strategy along the time gradient since abandon-

Table 2. Summary of the lognormal regression equations developed for richness and for cover in relation to time since abandonment for allspecies �total� and for the different life forms and dispersal strategies �see Figure 1 to 5�. a, b, x0 and y0 are the regression parameters of theequation y � a exp�-0.5 �ln�x/x0�/b�2� � y0. Cover of woody plants linear regression is y � y0 � a x. Only significant regressions areincluded �parameter mean � standard error�.

a X0 b Y0 R2 p

Life formsCover

Annuals 64.93 � 4.51 1.55 � 0.25 1.06 � 0.14 – 0.632 � 0.0001Forbs 38.66 � 4.96 5.71 � 0.37 3.17 � 0.44 10.40 � 2.54 0.424 � 0.0001Grasses 53.31 � 4.42 18.32 � 0.75 0.63 � 0.05 5.22 � 0.90 0.615 � 0.0001Woodies 1.34 � 0.07 – – 2.59 � 1.73 0.885 � 0.0001

RichnessTotal 23.10 � 0.8 10.52 � 1.17 1.99 � 0.17 – 0.356 � 0.0001Annuals 11.34 � 0.73 2.97 � 0.16 0.89 � 0.06 – 0.669 � 0.0001Forbs 9.10 � 0.42 6.74 � 0.58 1.37 � 0.09 – 0.448 � 0.0001Grasses 3.38 � 0.20 28.12 � 8.40 1.68 � 0.31 – 0.384 � 0.0001Woodies 12.72 � 1.60 68.65 � 46.7 1.74 � 0.44 – 0.571 � 0.0001

Dispersal modeCover

Anemochory 55.29 � 2.94 10.38 � 1.57 1.84 � 0.22 – 0.203 � 0.0001Ectozoochory 26.83 � 3.31 1.20 � 0.27 2.26 � 0.53 – 0.320 � 0.0001

RichnessAnemochory 6.56 � 0.73 6.0 � 0.52 0.68 � 0.11 5.32 � 0.53 0.488 � 0.0001Barochory 12.05 � 1.02 65.05 � 57.16 2.80 � 0.69 – 0.412 � 0.0001Ectozoochory 3.49 � 0.27 3.11 � 0.39 1.30 � 0.14 – 0.423 � 0.0001

262

ment was only significant for anemochorous andectozoochorous species �Figure 4; Table 2�. Ectozoo-chorous species only appeared at early stages of suc-

cession � � 15 years after abandonment�. Anemo-chorous species peaked at approximately intermediatestages, and barochory occurred at all stages without

Figure 4. Cover �%� of plant species with different dispersal strategy along successional gradient. Fitted lines are the significant non-linearregressions �see Table 2 for more details�.

263

any clear pattern. Endozoochory showed no statisti-cal pattern either but tended to increase with time.

Species richness of the different dispersal strategiesshowed different patterns along the abandonment agegradient �Figure 5, Table 2�. Anemochory and ecto-

zoochory species richness peaked at the early stagesof the analysed period � � 15 years�. Barochory spe-cies richness increased during the first 20 years andthen was maintained. Mirmecochory and endozoo-chory did not show any clear pattern. The number of

Figure 5. Richness �on 100 m2-plots� of plant species with different dispersal strategy along successional gradient. Fitted lines are the sig-nificant non-linear regressions �see Table 2 for more details�.

264

dispersal types did not show a clear trend along theland abandonment gradient.

Discussion

Our results suggest that total plant cover in semi-aridMediterranean systems is not dependent on abandon-ment age, and high cover values can be found at anytime following abandonment. The results also suggestthat, even in semi-arid conditions �annual rainfall ca.300 mm�, total species richness shows a peak at earlystages of the chronosequence �Fig 1�. The early peakpattern in richness appears to be in conflict with thedecreasing pattern of species richness found by De-bussche et al. �1996� in mesic Mediterranean condi-tions. However, their study did not include old-fieldsyounger than ca. 20 years, and so they were not ableto detect the pattern at early abandonment stages. Inthis sense, our results complement those by De-bussche et al. �1996�. In fact, the original work byEscarré et al. �1983� showed a peak in richness atearly stages � � 20 years�. This pattern may be attrib-utable to the importance of immigration processesduring the early stages of succession and the domi-nance of extinction processes after ca. 15 years sinceabandonment. The peak in richness described byTatoni and Roche �1994� for an abandonment gradi-ent on terraces in Provence �Mediterranean southernFrance� showed maximum values at intermediatestages on an ordination gradient, but without any ref-erence to age since abandonment.

A comparison of the trends in species richness in-dicates that the richness peak is the result of the co-existence of annual and biennial species withperennial forbs, some grasses, and woody plants dur-ing the first decades of the chronosequence �Figure 2,Figure 3�. Although all four life forms coexist, thereis a tendency towards life-form replacement, asshown by the displacement of the peak �annuals,forbs, grasses, and woody species, consecutively�when looking at the different life forms �Figure 2�.Annuals and forbs showed more markedly unimodalresponses than perennial grasses and woody species,indicating that the transition between life forms ismore abrupt between pioneers and the rest, as pointedout by Brown �1992� when comparing different suc-cession case studies. Seed bank composition plays animportant role in the peak of annuals and forbs atearly stages of abandonment because many of these

species are ruderals or weeds present as seeds in theagricultural fallows �Baskin and Baskin 1998�.

The main difference observed between this patternand the classic replacement patterns suggested formesic Mediterranean ecosystems �Houssard et al.1980; Escarré et al. 1983� has to do with the time lagin which these changes occur. In our case, about 45%of total woody species richness is reached 10 yearsafter abandonment �Figure 3�, while in mesic Medi-terranean conditions �southern France�, woody spe-cies reached less than 20% of total richness at thistime �Escarré et al. 1983�. This pattern could be par-tially attributed to an early colonisation of some shrubspecies �e.g., Rhamnus lycioides� through the facili-tation of bird-dispersed seeds by cultivated trees act-ing as perches �McDonell and Stiles 1983; Ne’emanand Izhaki 1996; Verdú and García-Fayos 1996,1998�. In fact, most �ca. 90%� of the endozoochorousspecies were woody, and they were present through-out the entire chronosequence without showing anyperiod of dominance; this pattern is similar to the onereported in Near-East vineyards �Ne’eman and Izhaki1996�. It is interesting that woody species seem toappear earlier in dry than in mesic conditions. Thismay be attributable to the lower competition forwoody species seedlings in drier conditions �highersite availability� than in mesic ones where a thickherbaceous layer is common �Keeley et al. 1981;Davis et al. 1998; Vilà and Sardans 1999�. Moreover,some old-field chronosequences in non-Mediterra-nean conditions �e.g., east-central Minnesota, USA,Lawson et al. 1999;� did not show any successionaltrend in the abundance of woody species.

The high species richness at early stages of thechronosequence could also be inferred from coexist-ence of different dispersal types �Hovestadt et al.2000�. The recently abandoned fields had moreanemochorous and ectozoochorous species than theearlier abandoned �older� fields, supporting the shiftin dominance due to differences in dispersal mode aspredicted by the ‘relay floristics’ model �Egler 1954;Myster 1993�. However, barochory, endozoochoryand mirmecochory were also present at the beginningof the chronosequence.

Several authors have found that wind is a moreimportant dispersal vector in early successionalstages, while dispersal by animals becomes importantonly at later successional stages �Houssard et al.1980; Hodgson and Grime 1990�. Our results inanemochory trends are consistent with these observa-tions. However, when analysing animal-dispersed

265

species in our study, we found that ectozoochorousspecies are mainly associated with young stages ofthe chronosequence. Endozoochorus �bird-dispersed�species are also present at early stages, but did notshow a clear trend along the abandonment time gra-dient. Previous observations in other Mediterraneanold-fields �Debussche et al. 1982; Verdú and García-Fayos 1998�, indicated that bird-dispersed specieswere important at early stages.

The results also support the idea that patterns re-lated to plant species interactions are more interpret-able when the species are segregated into differentfunctional groups �Pausas and Austin 2001�. For ex-ample, both cover and richness showed differentialpatterns and improved prediction �higher explainedvariance� when species were separated by life form.Although life forms are crude functional types, theyhave been suggested as a first approximation for clas-sifying species by function �Lavorel et al. 1997�. Theincrease in predictability achieved by subdividing to-tal species richness into different plant types has alsobeen observed in different ecosystems �Nilsson et al.1989; Moore and Keddy 1989; Pausas 1994; Pausasand Carreras 1995; Austin et al. 1996; Leathwick etal. 1998; Pausas et al. 1999�. Functional classifica-tions like life forms can be useful tools for predictionand for inter-regional comparisons �Pausas andLavorel 2003� of richness and cover trends alongsuccessional gradients.

Although we found clear patterns of richness andcover in relation to abandonment age, the modelscould be improved by considering the different con-ditions of the old-fields. Small differences in environ-mental conditions �e.g., altitude, soil type�, type andtree-density of the old crops �e.g., Ne’eman andIzhaki 1996; Bonet 2004�, and different disturbanceregimes �mainly grazing�, should be considered for abetter understanding of the mechanisms driving pat-terns through abandonment age and succession �Bo-net, 2004�.

Acknowledgements

We gratefully acknowledge some field data providedby Antonio de la Torre and María Cremades. Earlyversions of the manuscript were improved byvaluable comments from S. Díaz, A. van der Valk andan anonymous reviewer. The research for this paperwas partially funded by the Aquadapt EU Project�EVK1-CT-2001-00104�, the Spanish Ministerio de

Ciencia y Tecnología �CICYT project REN2000/0529HID�, and the Valencia Government, Conselle-ria de Cultura, Educació i Ciència, GeneralitatValenciana �project GV97-RN14-4�. CEAM isfunded by Generalitat Valenciana and Bancaixa.

Appendix

Table A1. Complete list of the species found on the plots. Raunki-aer’s life form: Ch � chamephyte; H � hemicryptophyte; P �phanerophyte; Th � therophyte. Considered life forms: A � an-nual or biennial; F � perennial forb; G � perennial grass; W �woody �d � dwarf scrub, s � shrub, t � tree�. Main dispersalstrategy: a � anemochory; b � barochory; e � ectozoochory; n �endozoochory; m � mirmecochory. Nomenclature follows Mateoand Crespo �1998�.

Species name Raunkiaer’slife form

Lifeform

Dispersalstrategy

Aegilops geniculata Th A aAizoon hispanicum Th A bAjuga chamaepitys Th A bAjuga iva Ch F bAllium ampeloprassum G F bAmaranthus blitoides Th A eAmaranthus muricatus Th A bAnacyclus clavatus Th A aAnacyclus valentinus Th A aAnagallis arvensis Th A eAnthyllis cytisoides P Wd bAnthyllis terniflora P Wd bArgyrolobium zannoni Ch F bAristolochia pistolochia G F bArtemisia barrelieri P Wd aArtemisia herba-alba P Wd aAsparagus horridus Ch Wd nAsperula aristata ssp. scabra H F aAsphodelus fistulosus G F bAsphodelus ramosus G F bAster squamatus Th A aAstragalus hispanicus Ch F bAtractylis cancellata Th A aAtractylis humilis Ch F aAvena barbata Th A aAvenula bromoides H G aBallota hirsuta H F eBeta vulgaris ssp martitima Th A bBrachypodium distachyon H G aBrachypodium phoenicoides H G bBrachypodium retusum H F aBrassica fruticulosa Th A bBromus diandrus Th A aBromus madritensis Th A aBromus rigidus Th A eBromus rubens Th A aBupleurum fruticescens P Wd bCalendula arvensis Th A e

266

Table A1. Continued.

Species name Raunkiaer’slife form

Lifeform

Dispersalstrategy

Cardaria draba Th A bCarduus bourgeanus Th A aCarduus tenuifolius Th A aCarex halleriana H G aCarrichtera annua Th A eCarthamus lanatus Th A aCentaurea aspera ssp steno-phylla

Ch F a

Centaurea calcitrapa H F aCentaurea melitensis H F aChenopodium album Th A bChenopodium murale Th A bChenopodium vulvaria Th A eChrozophora tinctoria Th A bCichorium intybus Th A aCirsium arvense Th A aCirsum vulgare H F aCistus albidus P Wd bCistus clusii P Wd bConvolvulus althaeoides H F bConvolvulus arvensis Th A bConvolvulus lanuginosus Ch F bConyza canadensis Th A aCoronilla minima ssp. lotoides Ch Wd bCrepis vesicaria H F aCuscuta epithymum Th F bCynodon dactylon H G aCynoglossum ternifolium H F bDactylis glomerata ssp his-panica

H G a

Daucus carota H F eDescurainia sophia Th A bDiplotaxis erucoides Th A bDittrichia viscosa P Wd aDorycnium pentaphyllum Ch Wd bEchium creticum ssp. coyn-cianum

H F b

Ephedra fragilis P Ws nErica multiflora P Ws bErodium cicutarium H F eErodium malacoides H F eEruca vesicaria Th A bEryngium campestre H F eEuphorbia exigua Th A mEuphorbia falcata Th A mEuphorbia helioscopia Th A mEuphorbia lagascae Th A mEuphorbia mariolensis Th A mEuphorbia segetalis Th A mEuphorbia serrata Ch F mFagonia cretica Ch F bFestuca capillifolia H G aFestuca valentina H G aFilago pyramidata Th A aFoeniculum vulgare ssp. pip-eritum

H F a

Table A1. Continued.

Species name Raunkiaer’slife form

Lifeform

Dispersalstrategy

Fumana ericoides Ch Wd bFumana hispidula Ch Wd bFumana laevis Ch Wd bFumana thymifolia Ch Wd bFumaria parviflora H F bGalactites tomentosa Th A aGalium fruticescens Th A bGalium verrucosum Th A bGenista scorpius P Wd bGlobularia alypum P Wd aHalogeton sativus Th A bHammada articulata P Wd bHaplophyllum linifolium Ch Wd bHelianthemum cinereum ssp.rotundifolium

Ch Wd b

Helianthemum syriacum Ch Wd bHelianthemum violaceum Ch Wd bHelianthemun hirtum Ch Wd bHelianthemun squamatum Ch Wd bHelichrysum decumbens Ch F aHelichrysum stoechas Ch Wd aHelictotrichon filifolium H G aHippocrepis ciliata Ch F eHordeum murinum ssp. lepori-num

Th A a

Hyparrhenia hirta H G aHyparrhenia sinaica H G aHypericum perforatum H F bIberis hegelmairei Th A aJuniperus oxycedrus P Ws nKochia scoparia Th A bKoeleria vallesiana H G aLamarckia aurea Th A aLavandula latifolia P Wd bLavatera arborea P Wd bLavatera cretica Th A bLeontodon logirostris H F aLeuzea confiera H F aLinum narbonense Ch F bLinum strictum Th A bLinum suffruticosum H F bLithodora fruticosa Ch Wd bLolium rigidum Th A aLophocloa cristata Th A aLygeum spartum H G aMalcolmia africana Th A bMalva hispanica Th A bMalva neglecta Th A bMalva parviflora Th A bMarrubium vulgare Ch F bMatthiola fruticulosa Ch F bMedicago arabica Th A eMedicago littoralis Th A eMedicago sativa Th A bMelilotus sulcata Th A eMercurialis tomentosa Ch Wd b

267

Table A1. Continued.

Species name Raunkiaer’slife form

Lifeform

Dispersalstrategy

Misopates orontium Th A aMoricandia arvensis Th A bMuscari neglectum G F bOnobrychis stenorhiza Ch Wd bOnonis fruticosa P Wd bOnonis minutissima Ch Wd bOnonis sicula Th A bOnopordum macracanthus H F aOphrys apifera G F aOphrys fusca G F aOrobanche sp G F aPallenis spinosa H F aPapaver rhoeas Th A aParonychia argentea Ch Wd aPhagnalon rupestre Ch Wd aPhagnalon saxatile Ch F aPhlomis lychnitis Ch Wd aPinus halepensis P Wt aPiptaterum miliaceum H G aPistacia lentiscus P Ws nPlantago afra Th A aPlantago albicans Ch F ePlantago coronopus Th A aPlantago lagopus Th A aPlantago sempervirens H F aPoa annua Th A aPolygala rupestris H F bPotentilla reptans H F bPsoralea bituminosa Ch Wd eQuercus coccifera P Ws nRapistrum rugosum Th A bReichardia intermedia Th A aReichardia tingitana Th A aReseda lutea H F aReseda phyteuma Ch F aReseda undata H F aRhamnus lycioides P Ws nRosmarinus offıcinalis P Wd bRubia peregrina H F nRuta angustifolia H F aSalsola genistoides P Ws aSalvia verbenaca Ch Wd bSanguisorba minor H F eSantolina chamaecyparissusssp. squarrosa

P Wd a

Scabiosa atropurpurea H F eSchismus barbatus Th A aScirpus holoschoenus H G aScorpiurus sulcatus Ch F eScorzonera angustifolia Th A aScorzonera laciniata H F aSedum sediforme Ch F bSenecio malacitanus Ch F aSenecio vulgaris Th A aSideritis angustifolia Ch Wd aSideritis leucantha Ch Wd b

Table A1. Continued.

Species name Raunkiaer’slife form

Lifeform

Dispersalstrategy

Silene mellifera Ch F bSilene vulgaris H F bSisymbrium irio Th A bSonchus asper Th A aSonchus oleraceus Th A aSonchus tenerrimus Th A aStaehelina dubia Ch Wd aStipa capensis Th A aStipa offneri H G aStipa parviflora H G aStipa tenacissima H G aTeucrium capitatum Ch Wd bTeucrium carolipauli Ch Wd bTeucrium homotrichum Ch Wd bTeucrium pseudochamaepitys Ch Wd bThymelaea argentata Ch Wd bThymelaea hirsuta P Wd bThymus moroderi Ch Wd bThymus vulgaris Ch Wd bUlex parviflorus P Wd bVicia peregrina H F bVitis vinifera P Wd nVulpia ciliata Th A a

References

Auclair A.N. and Goff F.G. 1971. Diversity relations of upland for-ests in the Western Great Lakes area. Am. Nat. 105: 499–528.

Austin M.P., Pausas J.G. and Nicholls A.O. 1996. Patterns of treespecies richness in relation to environment in southeastern NewSouth Wales, Australia. Aust. J. Ecol. 21: 154–164.

Barber A., Cabrera M.R. and Guardiola I. 1997. Sobre la culturade l´espart al territori valencià. Fundación Bancaixa, Valencia,España.

Baskin C.C. and Baskin J.M. 1998. Seeds: ecology, biogeography,and evolution of dormancy and germination. Academic Press,San Diego, London. 666 pp.

Bazzaz F.A. 1975. Plant species diversity in old-field successionalecosystems in southern Illinois. Ecology 56: 485–488.

Bellot J., Bonet A., Sánchez J.R. and Chirino E. 2001. Likely ef-fects of land use changes on the runoff and aquifer recharge insemiarid landscape using a hydrological model. Landscape andUrban Planning 55: 41–53.

Bonet A. 1997. Efectos del abandono de los cultivos sobre la veg-etación en la cuenca del Alt Llobregat �Barcelona�. Relación confactores ambientales y de usos del suelo. Ecología 11: 91–104.

Bonet A. 2004. Secondary succession on semi-arid Mediterraneanold-fields in South-eastern Spain: insights for conservation andrestoration of degraded lands. Journal of Arid Environments 756:213–233.

Bonet A., Peña J., Bellot J., Cremades M. and Sánchez J.R. 2001.Changing vegetation patterns in semi-arid Spain. In: VillacampaY., Brebbia C.A. and Usó J.L. �eds�, Ecosystems and Sustain-

268

able Development III, Advances in Ecological Studies 10, 377–385. Wit Press, Southampton, UK.

Brown V.K. 1992. Plant succession and life history strategy. TrendsEcol. Evol. 7: 143–144.

Christensen N.L. and Peet R.K. 1984. Convergence during second-ary forest succession. J. Ecol. 72: 25–36.

Davis M.A., Wrage K.J. and Reich P.B. 1998. Competition betweentree seedlings and herbaceous vegetation: support a theory of re-source supply and demand. Journal of Ecology 86: 652–661.

Debussche M.; Escarré J. and Lepart J. 1982. Ornitochory and plantsuccession in Mediterranean orchards. Vegetatio 48: 255–266.

Debussche M., Escarré J., Lepart J., Houssard C. and Lavorel S.1996. Changes in Mediterranean plant succession: old-fields re-vised. Journal of Vegetation Science 7: 519–526.

Draper N.R. and Smith H. 1981. Applied Regression Analysis. 2nded. John Wiley and Sons, New York, USA.

During H.J. and Willems J.H. 1984. Diversity models applied to achalk grassland. Vegetatio 57: 103–114.

Egler F.E. 1954. Vegetation science concepts I. Initial floristiccomposition. A factor in old-field vegetation development. Veg-etatio 4: 412–417.

Escarré J., Houssard C. and Debussche M. 1983. Évolution de lavégétation et du sol après abandon cultural en région méditer-ranéenne: étude de succession dans le Garrigues du Montpel-liérais �France�. Acta Oecol. 4: 221–239.

Grime J.P. 1973. Competitive exclusion in herbaceous vegetation.Nature 242: 344–347.

Hodgson J.G. and Grime J.P. 1990. The role of dispersal mecha-nisms, regenerative strategies and seed banks in the vegetationdynamics of the British landscape. In: Bunce R.G.H. andHoward D.C. �eds�, Species Dispersal in Agricultural Habitats.Belhaven Press, London, UK, pp. 65-81.

Houssard C., Escarré J. and Romane F. 1980. Development of spe-cies diversity in some Mediterranean plant communities.Vegetatio 43: 59–72.

Hovestadt T., Poethke H.J. and Messner S. 2000. Variability in dis-persal distances generates typical successional patterns: a simplesimulation model. Oikos 90: 612–619.

Huston M.A. 1979. A general hypothesis of species diversity. Am.Nat. 113: 81–101.

Huston M.A. 1994. Biological Diversity. The Coexistence of Spe-cies on Changing Landscapes. Cambridge University Press,Cambridge, UK. 681 pp.

Keeley S., Keeley J.E., Hutchinson S.M. and Johnson A.W. 1981.Postfire succession of the herbaceous flora in southern Califor-nia chaparral. Ecology 62: 1608–1621.

Lavorel S., Lepart J., Debussche M., Lebreton J.D. and Beffy J.L.1994. Small scale disturbances and the maintenance of speciesdiversity in Mediterranean old fields. Oikos 70: 455–473.

Lavorel S., McIntyre S., Landsberg J.J. and Forbes T.D.A. 1997.Plant functional classifications: from general groups to specificgroups based on response to disturbance. Trends Ecol. Evol. 12:474–478.

Lawson D., Inouye R.S., Huntly N. and Carson W.P. 1999. Patternsof woody plant abundance, recruitment, mortality, and growth ina 65-year chronosequence of old-fields. Plant Ecology 145:267–279.

Lubchenco J. 1978. Plant species diversity in a marine intertidalcommunity: importance of herbivore food preference and algalcompetitive abilities. Am. Nat. 112: 23–39.

Maestre F.T., Bautista S. and Cortina J. 2003. Positive, negative,and net effects in grass-shrub interactions in Mediterraneansemiarid grasslands. Ecology 83, in press.

Margaris N.S., Koutsidou E. and Gioura Ch. 1996. Changes in tra-ditional Mediterranean land-use systems. In: Brandt J. andThornes J.B. �eds�, Mediterranean desertification and land use,29–42. Whiley and Sons, Chischester, UK.

Martínez-Fernández J., López-Bermúdez F., Martínez-Fernández J.and Romero-Díaz A. 1995. Land use and soil-vegetationrelationships in a Mediterranean ecosystem: El Ardal, Murcia,Spain. Catena 25: 153–167.

Mateo G. and Crespo M.B. 1998. Manual para la determinación dela flora valenciana. Ed. Monogr. Flora Montiber 3, Valencia, Es-paña. 483 pp.

McDonnell M.J. and Stiles E.W. 1983. The structural complexityof old field vegetation and their recruitment of bird-dispersedplant species. Oecologia 56: 109–116.

Montalvo J., Casado M.A., Levassor C. and Pineda F.D. 1995. Thestrategies of ecological succession: theoretical insights from ex-perimentally disturbed ecosystems. In: Bellan D., Bonin G. andEmig C. �eds�, Functioning and dynamics of natural andperturbed ecosystems. Lavoisier Publs., Intercept Ltd., Paris,France. pp.231-261.

Moore D.R.J. and Keddy P.A. 1989. The relationship between spe-cies richness and standing crop in wetlands: the importance ofscale. Vegetatio 79: 99–106.

Molinier R. and Müller P. 1938. La dissémination des especesvégétales. Comunication Sigma 94, Montpellier, France.

Myster R.W. 1993. Tree invasion and establishment in old-fields atHutchenson Memorial Forest. Bot. Rev. 59: 251–272.

Ne’eman G. and Izhaki I. 1996. Colonization in abandoned EastMediterranean vineyard. J. Veg. Sci. 7: 465–472.

Noy-Meir I. 1973. Desert ecosystems: environment and producers.Ann. Rev. Ecol. Syst. 4: 25–51.

Pausas J.G. 1994. Species richness patterns in the understorey ofPyrenean Pinus sylvestris forest. Journal of Vegetation Science5: 517–524.

Pausas J.G. in press. Changes in fire and climate in the eastern Ibe-rian Peninsula �Mediterranean basin�. Climatic Change.

Pausas J.G. and Carreras J. 1995. The effect of bedrock type, tem-perature and moisture on species richness of Pyrenean Scots pine�Pinus sylvestris L.� forests. Vegetatio 116: 85–92.

Pausas J.G. and Vallejo V.R. 1999. The role of fire in EuropeanMediterranean ecosystems. In: Chuvieco E. �ed.�, Remote sens-ing of large wildfires in the European Mediterranean basin. pp.3-16. Springer-Verlag, Berlin, Germany.

Pausas J.G. and Austin M.P. 2001. Patterns of plant species rich-ness in relation to different environments: An appraisal. Journalof Vegetation Science 12: 153–166.

Pausas J.G. and Lavorel S. 2003. A hierarchical deductive approachfor functional types in disturbed ecosystems. Journal of Vegeta-tion Science 14 �3�: 409–416.

Pausas J.G., Carbó E., Caturla R.N., Gil J.M. and Vallejo V.R.1999. Post-fire regeneration patterns in the Eastern Iberian Pen-insula. Acta Oecol. 20: 499–508.

Peco B., Sánchez G. Casado M.A. and Pineda F.D. 1991. Dina-mismo de la diversidad y estructura espacial de pastizales medi-terráneos periódicamente perturbados. In: Pineda F.D., CasadoM.A., de Miguel J.M. and Montalvo J. �eds�, Diversidadbiológica/Biological diversity. F. Areces-ADENA-SCOPE,Madrid, Italy. pp. 215-218.

269

Peet R.K. 1978. Forest vegetation of the Colorado Front Range:patterns of species diversity. Vegetatio 37: 65–78.

Peet R.K. and Christensen N.L. 1988. Changes in species diversityduring secondary forest succession on the North Carolina Pied-mont. In: During H.J., Werger M.J.A. and Willems J.H. �eds�,Diversity and pattern in plant communities. pp. 233-245. SPBAcademic Publishing, The Hague, The Netherlands.

Pugnaire F.I. and Luque M.T. 2001. Changes in plant interactionsalong a gradient of environmental stress. Oikos 93: 42–49.

Rego F.C. 1992. Land use changes and wildfires. In: Teller A.,Mathy P., Jeffers J.N.R. �eds�, Response of forest fires to envi-ronmental change. Elsevier, London, UK, 367–373.

Scales L. E. 1985. Introduction to Non-linear Optimization.Springer-Verlag, New York, New York, USA.

Schelesinger W.H., Reynolds J.F., Cunningham G.L., HuennekeL.F., Jarrell W.M., Virginia R.A. and Whitford W.G. 1990. Bio-logical feedbacks in global desertification. Science 247: 1043–1048.

Sokal R.R. and Rohlf F.J. 1995. Biometry, 3rd. ed. W. H. Freemanand Co., New York, New York, USA.

Tatoni T. and Roche P. 1994. Comparison of old-field and forestrevegetation dynamics in Provence. Journal of Vegetation Sci-ence 5: 295–302.

Tilman D. 1982. Resource competition and community structure.Princeton University Press, Princeton, New Jersey, USA.

Tilman D. 1985. The resource-ratio hypothesis of plant succession.Am. Nat. 125: 827–852.

van der Pijl L. 1972. Principles of Dispersal on Higher Plants.Springer-Verlag, Berlin, Germany.

van der Valk A. 1992. Establishment, colonization and persistence.In: Glenn-Lewin D.C., Peet R.K. and Veblen T.T. �eds�, Plantsuccession: Theory and prediction. Champman and Hall, Lon-don, UK. pp. 60-102.

Verdú M. and García-Fayos P. 1996. Nucleation processes in aMediterranean bird-dispersed plant. Funct. Ecol. 10: 275–280.

Verdú M. and García-Fayos P. 1998. Old-field colonization byDaphne gnidium: seedling distribution and spatial dependence atdifferent scales. Journal of Vegetation Science 9: 713–718.

Vilà M. and Sardans J. 1999. Plant competition in mediterranean-type vegetation. Journal of Vegetation Science 10: 281–294.

Wheeler B.D. and Giller K.E. 1982. Species richness of herbaceousfen vegetation in Broadland, Norfolk in relation to the quantityof above-ground plant material. J. Ecol. 70: 179–200.

Willson M.F. 1993. Dispersal mode, seed shadows, and coloniza-tion patterns. Vegetatio 107/108: 261–280.

Wilson S.D. and Keddy P.A. 1988. Species richness, survivorship,and biomass accumulation along an environmental gradient. Oi-kos 53: 375–380.

Wisheu I.C. and Keddy P.A. 1989. Species richness-standing croprelationships along four lakeshore gradients: constraints on thegeneral model. Can. J. Bot. 67: 1609–1617.

Wright S.J. 1992. Seasonal drought, soil fertility and species den-sity of tropical forest plant communities. Trends Ecol. Evol. 7:260–263.

270